StarRocks
activeby StarRocks (Linux Foundation; CelerData) · since 2020
MPP OLAP database with a fast vectorized engine, real-time upserts and native lakehouse queries.
Overview
StarRocks is an open-source MPP analytical database (a 2020 fork of Apache Doris, now a Linux Foundation project) with a C++ vectorized engine and a cost-based optimizer that handles multi-table joins well. It serves both its own primary-key tables, which support real-time upserts from Kafka or Flink, and external Iceberg, Hudi, Delta and Hive tables directly, with async materialized views that transparently rewrite queries. Shared-data mode stores data on object storage with local caching. It is a strong pick for joins-heavy, high-concurrency analytics; the frontend/backend architecture and MySQL-flavoured SQL take some getting used to.
Strengths & trade-offs
Strengths
- +Cost-based optimizer handles star and snowflake joins at interactive speed
- +Primary-key tables support real-time upserts and deletes
- +Queries Iceberg, Hudi, Delta and Hive tables in place with caching
- +Async materialized views with automatic query rewrite
- +MySQL protocol works with existing BI tools and clients
Trade-offs
- −FE/BE/CN architecture adds operational components
- −MySQL-flavoured SQL differs from ANSI/Postgres in places
- −Smaller global community than ClickHouse; much activity is China-based
- −Frequent major releases require upgrade planning
Great for
- +Real-time dashboards with joins
- +Lakehouse query acceleration over Iceberg
- +Customer-facing analytics
- +Replacing pre-aggregated cubes
Think twice for
- −OLTP
- −Tiny single-node workloads
- −Document or key-value access patterns
Capabilities
Native or official first-party support · 13 of 25 tracked
Scale & consistency
- Comfortable data size
- Large · ≤ 1 PB
- Horizontal scaling
- Native
- Multi-region
- No
- Transactions
- Limited ACID
- Consistency
- Strong
- Schema
- Strict
- Architecture
- Distributed, Shared storage
- Query languages
- SQL (MySQL dialect)
Frontend (FE) and backend/compute nodes scale out; shared-data mode puts storage on S3 with elastic compute.
Deployment
Ecosystem
Alternatives to StarRocks
Real-time MPP analytical database with MySQL compatibility, search indexes and lakehouse access.
starrocks vs apache-doris →Open-source columnar OLAP database known for very fast aggregations over billions of rows.
starrocks vs clickhouse →Real-time OLAP store from LinkedIn for user-facing analytics at very high query rates.
starrocks vs apache-pinot →Real-time analytics database for sub-second slice-and-dice over streaming event data.
starrocks vs apache-druid →Fast distributed SQL query engine for federated analytics across lakes and databases.
starrocks vs trino →Commercial cloud and enterprise distribution of Apache Doris from its core maintainers.
starrocks vs velodb →StarRocks FAQ
What is StarRocks?
StarRocks is a columnar / OLAP database by StarRocks (Linux Foundation; CelerData), first released in 2020. StarRocks is an open-source MPP analytical database (a 2020 fork of Apache Doris, now a Linux Foundation project) with a C++ vectorized engine and a cost-based optimizer that handles multi-table joins well. It serves both its own primary-key tables, which support real-time upserts from Kafka or Flink, and external Iceberg, Hudi, Delta and Hive tables directly, with async materialized views that transparently rewrite queries. Shared-data mode stores data on object storage with local caching. It is a strong pick for joins-heavy, high-concurrency analytics; the frontend/backend architecture and MySQL-flavoured SQL take some getting used to.
Is StarRocks open source?
Yes. StarRocks is open source. License: Apache License 2.0.
Does StarRocks support ACID transactions?
Partly. StarRocks offers limited ACID guarantees (for example, single-record or single-partition transactions), with strong consistency.
Does StarRocks scale horizontally?
StarRocks scales horizontally out of the box, with built-in sharding or partitioning. It is comfortable at large scale (≤ 1 PB).
How can I run StarRocks?
You can self-host it. Managed offerings include CelerData Cloud.
What is StarRocks best for?
StarRocks is a strong fit for real-time dashboards with joins, lakehouse query acceleration over Iceberg and customer-facing analytics. Think twice for OLTP.
What are the best alternatives to StarRocks?
Common alternatives to StarRocks include Apache Doris, ClickHouse, Apache Pinot, Apache Druid and Trino. The right choice depends on your workload, scale and hosting; WhichDB's smart match ranks them against your requirements.
How popular is StarRocks?
StarRocks ranks #210 of 844 databases on the WhichDB Score (15.6/100), and #22 among columnar / OLAP databases. The score combines public adoption signals such as package downloads, GitHub activity, job posts and developer discussion.